Bibliotheca Arnamagnæana - 01.06.1959, Síða 78
64
this branch. But some of the French sources dealt with events both before
and after the battie of Roncevaux, and in such cases he had to divide these
stories in two.
The editor had before him the following tales (translated under his
supervision, or already existing as separate sagas):
a) A Vie romancée de Charlemagne. I fully agree with M. Aebischer
about the nature of this branch, and with his (and Storm’s) opinion that
the source was written in French, not Latin. M. Aebischer is also right in
thinking that what corresponds to branch X in D originally belonged to
this branch, and his assumption that the Vie romancée once contained a
brief description of the battie of Roncevaux is certainly correct.
b) A chanson de geste resembling the Chevalerie Ogier de Danemarche.
The first part describes Ogier’s youth, and its place had to be at the be-
ginning of the saga, after the account of Charlemagne’s early years. But
the last part of this poem shows us Ogier as an old man, and since Roland
does not take part in the fighting, he must obviously be dead; this part
therefore had to be placed after branch VIII. As for the end of branch III,
which is missing in Aa, it is certainly derived from the same French source
as the rest of this branch, since it is found in Bb and D. I have previously
(pp. 53-54) tried to explain the absence of these chapters in Aa by assum-
ing that a leaf had been lost in the MS from which A and a are derived.
Another explanation is perhaps possible: the editor of the original Kms,
when he divided the story of Ogier le Danois into two parts, may have
placed chapters 47—54 of branch III with the “Wars against Amarus and
Maskabret” after branch VIII, because the heroes in both cases were
Charlot and Ogier. This was obviously wrong, and the editors of Bb (the
Bb version, not the writer of the MS on which our MSS B and b are
ultimately based) and D, who both had some critical sense, have in-
dependently moved the chapters to their correct place. But I admit that
this explanation is not completely satisfactory.
c) The Chanson d’AspremonV and the Pseudo-Turpin. He identified
the Agolandus of Turpin with the Aigolant of the French poem, and com-
bined the two, without bothering to get rid of the contradictions. He in-
cluded most of Pseudo-Turpin, omitting only the account of the battie of
Roncevaux, which was described at greater length in the Chanson de
81 R. van Waard: Études sur l’origine et la formation de la Chanson d’Aspremont,
pp. 189-215.